Q&A: CBR talks to Magnitogorsk Iron and Steel Works about their deployment of Yandex Data Factory technology.
The use of data within an organisation to improve elements of the business such as the supply chain, improve decision making, and to make cost savings, is becoming more widely accepted as being vital.
It is vital in respect to the business remaining competitive, vital to remaining relevant, and vital to the future of the business.
One of the industries that has been looking significantly at the use of its data is the manufacturing industry, and stepping back one level to the steel industry.
Magnitogorsk Iron and Steel Works (MMK), is the third largest steel company in Russia with a revenue of $9.3bn.
Established in 1870, the company has taken to using machine learning technology from Yandex Data Factory to creative a competitive advantage that will see it being competitive for years to come.
CBR spoke to Sergey A. Sulimov, Deputy CEO for Finance and Economy, MMK OJSC:
JN: Tell me a little about your IT stack, what are the core technologies you use?
Sulimov: “If we are talking about “big data”, then machine learning is the first technology that is used at MMK in the field of automatic data processing. In addition to this we use Cloudera Hadoop as a data storage system.”
JN: Before using machine learning and big data analytics from Yandex Data Factory what were you using?
Sulimov : “We didn’t use any data processing technology. The data was accumulated and used for recording operational activity and generating quick reports. Analytical reports were designed through the use of separate technical specifications.”
JN: How exactly are you using the Yandex Data Factory’s machine learning solution now?
Sulimov: “Yandex Data Factory’s data scientists developed a machine learning service that analyses given input parameters to produce recommendations on the optimal quantity of materials used for steel production.”
JN: Is this a company-wide deployment?
Sulimov: “It’s a MES-level (manufacturing execution systems) implementation.”
JN: What was the process in getting it set up? Were there any challenges?
Sulimov: “The actual development of the service was done by Yandex Data Factory specialists. The role of MMK’s IT specialists was to prepare the historical data for further analysis. The main challenge was the structure and the quality of data generated by the evolving MES-systems.
“With the development of the MES-system more and more data was generated on the same object, and with the change of technology the ranges of values also changed. Quality control (validation of this data) has taken a significant amount of time.”
JN: How exactly is the data being used to save money?
Sulimov: “We use accumulated data on metal stock spent overall (in particular the ferroalloys) and actual values of chemical elements contained in steel.”
JN: Where do you store the data once it has been analysed? Has machine learning helped you to decrease data storage expenditure?
Sulimov: “The main part of data was stored and continues to be stored in a classic relational database. The data that is prepared and formalised for our joint project with Yandex Data Factory, is stored in Hadoop.
“Now we’re in the process of developing an integration system that allows you to store all the technical information on the project in Hadoop, including the opportunity for its further analysis.”
JN: Do you think machine learning will be a key competitive differentiator for you?
Sulimov: “Preliminary calculations show that this machine learning service will let us reduce the production cost of steel. This will be a significant competitive advantage for MMK.”